import json
import csv
from pathlib import Path
from datetime import datetime

# --- CONFIGURATION ---
BASE_PATH = 'C:/temp/discord/'
OUTPUT_FILENAME = 'output_timestamps_clean.csv'

def standardize_timestamp(ts_string):
    """Converts varying timestamp formats into a standard YYYY-MM-DD HH:MM:SS format."""
    if not ts_string:
        return None
        
    # Format 1: "02.06.2021 06:26"
    try:
        dt = datetime.strptime(ts_string, "%d.%m.%Y %H:%M")
        return dt.strftime("%Y-%m-%d %H:%M:%S")
    except ValueError:
        pass

    # Format 2: "2021-02-17T20:35:29.083Z" (ISO 8601)
    try:
        dt = datetime.strptime(ts_string.replace("\"","")[:19].replace('T', ' '), "%Y-%m-%d %H:%M:%S")
        
        #dt = datetime.fromisoformat(ts_string[:17].replace('T', ' '))
        return dt.strftime("%Y-%m-%d %H:%M:%S")
    except ValueError:
        pass
        
    # Fallback: return original string if it matches neither format
    return ts_string

def process_file(file_path, writer):
    """Attempts to read a file as standard JSON, falls back to JSON Lines if needed."""
    with open(file_path, 'r', encoding='utf-8') as f:
        try:
            # Attempt 1: Standard JSON
            data = json.load(f)
            records = data if isinstance(data, list) else [data]
            
            for entry in records:
                if isinstance(entry, dict):
                    # Check both capitalization variants just in case
                    raw_ts = entry.get('Timestamp') or entry.get('timestamp')
                    if raw_ts:
                        clean_ts = standardize_timestamp(raw_ts)
                        writer.writerow([clean_ts, file_path.name])
            return 
            
        except json.JSONDecodeError:
            pass 

        # Attempt 2: JSON Lines
        f.seek(0) 
        
        for line in f:
            line = line.strip()
            if not line:
                continue
                
            try:
                entry = json.loads(line)
                records = entry if isinstance(entry, list) else [entry]
                
                for rec in records:
                    if isinstance(rec, dict):
                        raw_ts = rec.get('Timestamp') or rec.get('timestamp')
                        if raw_ts:
                            clean_ts = standardize_timestamp(raw_ts)
                            writer.writerow([clean_ts, file_path.name])
                            
            except json.JSONDecodeError:
                pass

def extract_and_format_timestamps():
    work_dir = Path(BASE_PATH)
    output_path = work_dir / OUTPUT_FILENAME
    
    with open(output_path, 'w', newline='', encoding='utf-8') as csv_out:
        writer = csv.writer(csv_out)
        writer.writerow(['Timestamp', 'Filename'])
        
        print(f"Scanning: {work_dir}")
        print(f"Output will be saved to: {output_path}")

        for file_path in work_dir.glob('**/*.json'):
            if file_path.name == OUTPUT_FILENAME:
                continue
                
            try:
                process_file(file_path, writer)
            except Exception as e:
                print(f"Error reading file {file_path.name}: {e}")

    print("Done!")

if __name__ == "__main__":
    extract_and_format_timestamps()